Our long-term objectives are to develop a validated animal and human laboratory model, and in turn, treatment strategies, based on understanding the environmental factors that precipitate aberrant behaviors in persons with mental retardation (MR). Both humans and animals show prolonged, counterproductive interruptions in behavior (pausing) when relatively rich conditions of positive reinforcement transition to relatively lean conditions of positive reinforcement. When an escape response is available, subjects escape during these rich-lean transitions, indicating that this context is aversive. Clinical research has established escape as a primary motive for aberrant behavior, but has not illuminated the processes that render normally benign activities aversive. Pausing as maladaptive escape may provide a functional analogue to aberrant behavior, and conditions that generate long pausing may predict aberrant behavior. The proposed research will test the utility of this conceptualization in predicting the occurrence of stereotyped and self-injurious behaviors in persons with MR. In keeping with the translational nature of the research program, studies will be conducted in three contexts that span laboratory, naturalistic, and clinical settings. Three laboratory studies are proposed for each of two CAB topographies (self injury/aggression and stereotypy). Each laboratory study will be replicated under more naturalistic procedures. We then propose to study the effects of rich-to-lean transitions under conditions involving tasks of direct clinical relevance to the subjects. This research strategy is designed to better understand the behavioral processes that may provide aversive stimulation for escape motivated CAB, and begin to use this knowledge for developing treatment strategies.